Dear Gunnar,
As always I'm impressed by your erudition and detail. Seriously! From now
on, I'd be grateful if you would write my posts for me.
Yes, you detailed the events of that period beautifully. Yes, the issue of
who exactly are 'design staff' is a problem. Currently the employment
statistics people don't seem to distinguish very much between designers and
sweeper uppers of the print room.
Your suggestion for AI of graphics for logos may be already at least partly
in place with things like smart draw tools which convert sketches into
regular aligned shapes, in much the same way that Garageband converts an out
of time music riff into a regularised form.
Two references from MIT review as you requested, the second is probably the
more interesting:
http://www.technologyreview.com/news/428437/can-creativity-be-automated/
http://www.technologyreview.com/featuredstory/515926/how-technology-is-destr
oying-jobs/
You might also enjoy
http://www.magisto.com/how-it-works
On the layers of abstractions stuff, apologies, I guess its not the way
people think on this list, I'll work out a good physical example to show
it better. I tend to design in abstraction language which is great for many
areas of design to get beyond thinking about the physical. (and looking at
the dynamics of design variety is a really useful napkin design tool).
Enjoyed your stories about the quemma and other novel punctuation. I'd
thought they were also the sorts of things in the realms of novel computer
creativity? I remember creating software in the 1970s to optimise the
design of a mixed material bridge design. It was a university competition
and I thought by inventing a computer optimisation we might clean up (if I
remember right we did). What was more amazing to me, was that software
identified several novel solutions for bridge designs that were clearly
effective and totally different to the model that had been the basis for the
optimisation software. That was an early example of computer-based
creativity that went beyond what was put into the computer.
More recently, an IBM Watson has been let loose on creative cooking. . .
http://www.ibm.com/smarterplanet/us/en/cognitivecooking/
http://www.research.ibm.com/software/IBMResearch/multimedia/Cognitive-Cookin
g-Fact-Sheet.pdf
This more follows the big data approach that Ken raised.
PS finding roles for your at-persand and at-percent symbols sounds like
exactly an ontology algorithm problem. Ask IBM to ask Watson - but turn off
the urban dictionary first!
Best wishes and thanks ,
Terry
---
Dr Terence Love
PhD(UWA), BA(Hons) Engin. PGCEd, FDRS, AMIMechE, MISI
Director,
Love Services Pty Ltd
PO Box 226, Quinns Rocks
Western Australia 6030
Tel: +61 (0)4 3497 5848
Fax:+61 (0)8 9305 7629
[log in to unmask]
--
-----Original Message-----
From: [log in to unmask]
[mailto:[log in to unmask]] On Behalf Of Gunnar Swanson
Sent: Friday, 9 May 2014 4:30 AM
To: PhD-Design - This list is for discussion of PhD studies and related
research in Design
Subject: Re: Why designers need maths
On May 8, 2014, at 8:59 AM, Terence Love <[log in to unmask]> wrote:
> An example, in the late 1990s, large graphic design firms cut their
> design staff by around 75% and at the same time were able to increase
> output. In that case, the reduction in staff and increase in outputs
> were mainly due to the increased productivity enabled by products from
> Quark Express, Macromedia and Adobe.
This doesn't track with my memory of the graphic design business (at least
in the US.) The cuts were earlier, those not attributable to economic
downturns were not as deep, and the nature of the cuts was not exactly as
you describe. Or at least not neatly so.
The first big wave of job destruction was at suppliers for design firms.
Typesetters and 'stat houses disappeared except for the few that converted
to computer graphics service bureaus and produced film from designers'
QuarkXPress, etc. files. (At that point, printers had big staff reductions
in the stripping department. Strippers--the people who created and put
together the various pieces of film to make plates--had been the most
skilled and best paid people in printing.) Designers inherited much of the
responsibility that formerly belonged to typesetters, 'stat houses, and
strippers, so the efficiencies of computerization affected designers'
outside expenses more than their employment.
This brings us to the question of what you mean by "design staff." The
computer took over much of the role of the production people at design
firms. Were they "design staff"? Maybe. Were they designers? It depends on
which firms and who is describing them. One could argue that what they were
called is more of a political question than anything else but it gets back
to the eternal problem on this list: What do we actually mean by "design,"
"designer," or "design staff"?
> other changes due to automation that were increasingly substituting
> instead of designers' designerly knowledge and expertise.
Yes, although initially, the opposite happened. Designers who had looked
like they really understood type complained that "the Mac isn't up to good
typography" because they never understood how much detail work was done by
their typographers. They now needed to make decisions about kerning and such
that they had previously not faced. As you imply, much of that was
subsequently taken over by better software.
> Designers' natural human
> reactions against automation of their roles
That was the weird thing. What I saw was just the opposite. When computer
systems started looking viable (in the pre-Mac days), most of the designers
I knew expressed fear and most of the production artists I knew were excited
at the labor saving opportunities. This despite the fact that it seemed
clear that the designers would go on designing and the production artists
would be unemployed.
> The software is capable of
> doing far more of the human design decisionmaking than designers and
> many other in the design industry have been aware.
Yes. To some extent. In typography, for instance, it makes decisions that
just plain weren't made before and routinely makes decisions that were only
made in rarified, high budget situations.
> and critique possible new designs. The limit of designers learning
> and attributes is only the limit of the number of designs a person can
> see in their lifetime and their sensitivity to them. This and the use
> of emotions and thinking provides the creative competence of designers.
Perhaps the biggest change for graphic design is that realistic prototypes
can be created extremely fast, allowing more iterations on tighter time
and/or money budgets.
> One reality is that quite small computer systems can now process more
> than humans. More importantly, by processing large amounts of data
> they can learn the intrinsic tacit properties of that information and
> make it available to a wide variety of other processes. Graphic design
> is relatively unusual in that it has codified much of its knowledge
> and this makes it easier for computers to extend faster into the
> arena of meaning and automating human processes in graphic design.
So it seems to me that the biggest net role for computers in graphic design
(as opposed to production for graphic design) is in ideation. We teach a
clear set of ideation procedures in the graphic design program here at East
Carolina University. It's hardly exclusive to us but it involves set actions
of finding keywords, expanding the list of keywords, identifying resonant
items, expanding on them, doing the same thing with images, then looking for
useful combinations by observing shape similarities and other opportunities
to combine ideas in meaningful ways. . .
I can imagine an AI system that delivers some of this very efficiently.
(Off your topic--it would be interesting to see how the possible elimination
of some of the drawing in the process affects the understanding on the part
of the designer. It might be romanticism on my part--although I think
not--to believe that gathering clippings doesn't promote the sort of
synthesis that drawing thumbnails does.) So there would be some loss that
might be completely outweighed by the volume and efficiency of an automated
visual thesaurus, etc.
> For the near future, this suggests many if not all of designers
> traditional design activities may become computer automated. This is
> already in place for automated generation of advertising images and
messages.
Can computers fairly easily replace the level of designer that sells
services on the $100 logo websites? I suspect so.
> MIT review describes other examples of automating creativity .
I'd love to here more about this.
> Yes, I can almost hear you say, 'What has that to do with designers
> learning mathematics?'
You have very good ears. I thought I was using my indoors voice and you're
half a word away.
> Currently, humans are slow at interpreting meaning, and computers are
> extremely slow at it. In contrast, in terms of rate of variety,
> humans are slow and computers are extremely fast.
I finally gave up but I used to spend time talking to anyone doing parallel
computing and such about an AI idea for logo design. You'd scan your sketch
and the computer would analyze it and come back with a series of questions:
This part looks like part of a circle/do you want it to be part of a circle?
This part looks to be about half the size of that part/do you want it to be
exactly half? This almost lines up horizontally with that/do you want it
lined up? The transition from this part of the line to that part of the line
is just barely abrupt/do you want it smooth? This seemingly important thing
is near the center/do you want it in the center? etc.
Another approach would be to not ask the questions but just produce
iterations and let the designer play a game of "warm". . . "cooler". . .
"hot". . . "that's it" with the auto-illustrator.
> If designers start using maths to manage abstractions of behaviours of
> designed objects, criteria and characteristics and then use maths to
> abstract the behaviours of those abstractions THEN there starts to
> emerge an advantage in favour of humans. This is because abstractions
> of the behaviour of abstractions about objects means the objects being
> addressed by humans (abstracts of abstractions) potentially represent
> large numbers of objects and hence massively increase the rate of
> variety. More importantly, the maths can be used to focus selection of
> elements towards optimal solutions - of advantage in competing against
brute force management of variety.
Sorry. I couldn't decode that. Remember, I'm a graphic designer. You have to
talk s l o w l y with us and in short sentences. This may be where you
explain the math thing. I'm still wondering.
> Remember
> if computers can learn to produce designs on the basis of best
> designs and best design practices of the best designers, it is going
> to be increasingly harder to stay ahead of the creative designs of the
computers.
Which, of course, leaves computers with the problem of how new stuff gets
produced to use as a model when the old stuff gets old. And graphic design
may be second only to fashion design in the how fast can old stuff get old*
race.
[*I invented a punctuation mark that I call a quemma. It's a question mark
with a comma replacing the dot. I think we need something with less of a
pause than a quemma to indicate the internal question in that last sentence.
The fate of the interrobang makes me think it's unlikely to happen but if
computers do take over writers' and editors' jobs, maybe I just have to
convince some programmers to sneak it in. I suspect that programmers won't
go for my at-persand and at-percent symbols. I like the way they look but I
can't figure out a need for them so an algorithm for use isn't exactly at
hand.]
Gunnar
Gunnar Swanson
East Carolina University
graphic design program
http://www.ecu.edu/cs-cfac/soad/graphic/index.cfm
[log in to unmask]
Gunnar Swanson Design Office
1901 East 6th Street
Greenville NC 27858
USA
http://www.gunnarswanson.com
[log in to unmask]
+1 252 258-7006
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]> Discussion of PhD
studies and related research in Design Subscribe or Unsubscribe at
https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|